• SPC presented the 2024 SWPO striped marlin stock assessment to SC20 and identified issues with fits to the index and size composition data.
• SC20 requested these issues be addressed and presented to SC21
• The US offered to assist in the redevelopment of the stock assessment and host SPC staff for a modelling workshop.

• Multifan-CL can only implement selectivity at age, and spline selectivity nodes can only be on age points
• Improving the resolution of the selectivity curve was accomplished by moving to a quarterly population dynamics model
• Unfortunately this results in seasonal recruitment dynamics which are inconsistent with our understanding of striped marlin biology

• Stock Synthesis can model selectivity at length
• Improves the resolution of the selectivity curve and fit to the composition data while maintaining annual population dynamics

Data

• Conflict between AU length/weight data
• Non-representative composition data (mixed flag fisheries)
• Representativeness of the CPUE index (large spike)
• Inclusion of TW size composition data

Model structure

• Model start year (1952 v. 1979)
• Data-weighting
• Refine growth parameterization (estimate internally using CAAL)
• Refine selectivities (time-blocks or seasonal selectivity)
• 2 sex model
• Dealing with large catch in 1954